Fundle.ai: First Party Data Platform for Loyalty India: End-to-End Journeys

Mapping the Customer Journey in Indian Retail

Before any data platform can be configured, operators must complete a journey map that is honest about the Indian omnichannel reality. The Indian shopper's path to purchase does not follow a clean funnel. It begins with awareness on Instagram Reels or a WhatsApp forward from a relative, moves to in-store browsing at a mall like Select CITYWALK or Nexus Seawoods, detours through a price-check on Flipkart, and often closes either in-store on impulse or via a brand's own app with a first-order discount. Post-purchase, the relationship is maintained — or more accurately, neglected — through generic broadcast SMS and occasional emailers.

A proper journey map for Indian retail should identify at minimum eight distinct touchpoints: (1) first-party digital acquisition (app install, web form, WhatsApp opt-in), (2) in-store visit detection (beacon, Wi-Fi probe, POS swipe), (3) transaction at POS — whether integrated with POSist, GoFrugal, Petpooja, or Wondersoft — (4) post-purchase NPS collection, (5) browse abandonment on the brand's own digital property, (6) tier-upgrade trigger communications, (7) referral or social sharing event, and (8) lapse or churn signal detection. Each of these touchpoints generates a different data type: transactional, behavioural, attitudinal, or contextual. A first party data platform must ingest all four types and reconcile them against a single customer identity.

The identity resolution problem is particularly acute in India because most consumers interact with a brand across multiple phone numbers (primary SIM, secondary SIM, family member's number), multiple devices, and inconsistent name spellings in regional languages. A platform that cannot resolve 'Priya Sharma' registered on a Delhi mall app with 'P Sharma' who transacted at a Bangalore Lifestyle store is not truly unified. The data model must be probabilistic where deterministic matching fails, and it must surface confidence scores so marketing teams can act accordingly.

The journey map also needs to account for the role of mall operators as a separate data party. When a consumer walks into a Phoenix Marketcity property and earns points from both the mall loyalty programme and an in-store Pantaloons programme, the data needs to flow in both directions with appropriate consent scoping. Operators who invest in this mapping exercise upfront consistently report a 30–40% reduction in data deduplication costs and a 2x improvement in campaign match rates within the first six months.

Harnessing First Party Data at Each Touchpoint

Capturing data is the easy part. The harder discipline is ensuring that every touchpoint generates a signal that is clean, consented, and connected to a durable customer identifier. Indian retail operators routinely underinvest in this infrastructure, which is why the average identified-transaction rate sits at just 23% even among brands with active loyalty programmes. The other 77% of transactions are anonymous spend — value that cannot be attributed, retained, or grown.

At POS integrations, the gap is often technical: legacy billing software from vendors like Wondersoft or older GoFrugal deployments lack real-time API hooks, so loyalty data is synced in batch overnight. By the time a campaign team identifies a high-value buyer, the customer is already two purchases into a competitor's programme. Modern integrations must be event-driven — every transaction fires a webhook that updates the member's profile within seconds, not hours. This is non-negotiable for AI-driven next-best-offer engines to function correctly.

Beyond transactions, behavioural signals from the brand's own digital properties are chronically underused. A customer who opens the FabIndia app three times in a week but does not transact is exhibiting a high-intent signal that should trigger a personalised bridge offer — say, a limited-time bonus point event on handloom kurtas based on their browse history. WebEngage and MoEngage have made progress on this kind of event-triggered automation, but they lack the loyalty-native data model that connects behavioural intent to points, tiers, and redemption mechanics. A dedicated first party data platform for loyalty India operators need must treat points currency, tier status, and product affinity as first-class data objects, not afterthoughts bolted onto a marketing automation layer.

Mall operators face an additional complexity: aggregating data across dozens of tenants without violating individual brand data agreements. The answer is a federated data model where the mall holds the visitor identity and the brand holds the purchase detail — and both share only the signals necessary for joint campaign activation, governed by a master data-sharing agreement. This architecture, when built correctly, enables a Select CITYWALK to show a shopper that they are ₹1,200 away from a Silver tier upgrade across all tenant spend — a genuinely motivating insight that no single brand could surface alone.

Role of AI in Enhancing Loyalty Experiences

Artificial intelligence applied to a first-party data estate is not a feature — it is a force multiplier. The difference between a loyalty programme that delivers 8% incremental revenue lift and one that delivers 22% is almost entirely explained by the quality of AI application on top of a clean data layer. Indian retail has largely seen AI used for basic personalisation: 'customers like you also bought' recommendation carousels and broadcast-segment email subject-line A/B tests. That is table stakes, not competitive advantage.

The more consequential applications are predictive and agentic. Predictive churn models trained on Indian retail data — where shopping seasonality is driven by Diwali, Eid, wedding season, and annual school-uniform cycles rather than Western Black Friday patterns — can identify at-risk members 45 to 60 days before their lapse becomes permanent. A model that knows a Cafe Coffee Day regular who normally visits 12 times a month has dropped to 4 visits in the current month can trigger a personalised reactivation offer — a free upgrade on their next beverage order — before the habit is broken. The cost of that offer is a fraction of the cost of re-acquiring the customer through paid media.

Agentic AI takes this further. Rather than a marketer configuring a campaign workflow manually, Fundle AI Agents autonomously monitor member behaviour, detect anomalies, draft personalised outreach, select the optimal channel (WhatsApp, push notification, in-app message, or email), and dispatch communications within guardrails set by the brand team. Fundle Agentic AI and Fundle AI Workflow together create a system where the loyalty programme is genuinely always-on — not dependent on a campaign calendar that a three-person CRM team can manage. A mall operator running 80 tenant brands cannot manually orchestrate personalised journeys for 2 lakh active members. Agentic AI is the only operationally viable answer.

Natural language querying of loyalty data is another emerging capability. A CMO at a Lifestyle or Pantaloons should be able to ask: 'Which of our Gold members in Tier 2 cities have not purchased in the accessories category in 90 days but have browsed it?' and receive an actionable audience segment in under 30 seconds — not a three-day BI ticket. This is where the Fundle AI Platform's conversational analytics layer creates a meaningful productivity advantage for retail marketing teams that are perpetually under-resourced.

Maintaining Privacy and Consent Throughout

India's Digital Personal Data Protection Act 2023 is not a distant regulatory threat. The Data Protection Board is being constituted, sector-specific rules are being drafted, and the reputational cost of a consent violation in an era of consumer WhatsApp groups and Twitter/X virality is measurable in lost footfall. Indian retail CMOs and CIOs who treat DPDP compliance as a legal department problem rather than a data architecture problem will face expensive retrofits — or worse, enforcement action — within the next 24 months.

A privacy-first loyalty platform India operators need must embed consent as a data object, not a checkbox. This means: (a) capturing the specific purpose for which data is collected at the moment of collection ('We will use your purchase history to personalise offers' is a valid purpose; 'marketing purposes' is not under DPDP), (b) storing consent records with timestamps and version references to the privacy notice in effect at the time, (c) providing members with a genuine, friction-free mechanism to withdraw consent or request data deletion without requiring them to call a customer care line, and (d) implementing automated re-consent workflows when a new data processing purpose is introduced — for example, when a mall operator decides to share transaction data with a fintech partner for co-branded credit card offers.

The consent architecture must also account for children. India has a large population of Gen Z shoppers who are under 18, and DPDP explicitly requires parental consent for minors. A loyalty programme that allows 16-year-olds to sign up via WhatsApp without a verified parental consent flow is non-compliant from day one. Platform-level controls, not manual process workarounds, are the only scalable solution.

Data minimisation is the other underappreciated principle. Many Indian loyalty programmes collect date of birth, anniversary date, income bracket, and residential pin code at sign-up — fields that have never been used in a single campaign because the data team does not have the tooling to act on them. Collecting data that you cannot use is not just wasteful; under DPDP it is potentially unlawful. A rigorous data audit before platform deployment — mapping every collected field to an active use case — typically results in a 35–50% reduction in the data collection form, which paradoxically increases sign-up conversion rates because the form is shorter and the value exchange is clearer.

How Fundle solves this

Vineet Narang founded Fundle on a conviction that Indian retail deserved a loyalty and engagement platform built for its specific complexity — not a Western SaaS product localised with an INR currency symbol. The Fundle AI Platform is an end-to-end system that addresses every layer of the challenge described in this article: data capture, identity resolution, AI activation, consent management, and cross-brand journey orchestration.

For brand retailers — whether a national fashion chain, a jewellery brand like Tanishq managing high-value members, or an optical retailer like Lenskart with a digital-first base — Fundle Brand Loyalty provides a turnkey first party data platform for loyalty India operators can deploy in weeks, not quarters. It includes real-time POS connectors for all major Indian billing systems, a WhatsApp-native member enrolment flow that achieves 40–60% higher sign-up rates than SMS OTP methods, and a dynamic segmentation engine that refreshes RFM scores and churn propensity daily. Fundle supports comprehensive loyalty journeys for 1.33 Cr+ members across India — a scale that generates the training data necessary for AI models to be genuinely predictive rather than theoretically promising.

For mall operators managing multi-tenant ecosystems, Fundle Mall Loyalty provides the federated data architecture described earlier: a visitor identity graph owned by the mall, with tenant transaction data flowing into brand-specific loyalty ledgers and joint-campaign activation governed by configurable data-sharing rules. The result is a mall-level engagement layer that no individual tenant could build independently — and that creates measurable footfall increments of 12–18% for participating properties in the first year based on current deployments.

The AI layer — Fundle AI Agents, Fundle Agentic AI, and Fundle AI Workflow — converts this data estate into always-on journey orchestration. Churn prediction, next-best-offer selection, tier-nudge timing, and referral campaign triggering are all handled autonomously within brand-defined guardrails. Marketing teams shift from campaign execution to strategy and creative — a reallocation of human effort that consistently surfaces as the top reported benefit among Fundle's operator clients. The Fundle AI Workflow engine also integrates with downstream platforms including MoEngage, WebEngage, and Xeno for brands that want to preserve existing channel investments while upgrading their data and AI backbone.

On privacy, Fundle's consent management module is built to DPDP Act 2023 specifications, including purpose-linked consent capture, automated re-consent workflows, a member-facing data transparency portal, and data deletion SLAs configurable to 72 hours. For CIOs managing data governance obligations, this is not a vendor promise — it is a contractual commitment backed by the platform's architecture, not a manual compliance process.